Rangers rank ahead of Celtic in Global club football index for the season ahead

Recoba

Recoba

Well-Known Member
Official Ticketer
Before anyone says it i know it means nothing and its just predicted stats

14 places ahead of them

Some of it seems ridiculous but its another reason for the Timothy's around the world to become even more paranoid.

Where Celtic and Rangers rank in Global Club Soccer index ahead of new season revealed
The highly-detailed index ranks 636 international clubs - but where does your side sit ahead of the new premiership campaign?


Rangers have come out top of the Scots in the latest Global Club Soccer Rankings published ahead of the new domestic campaign.


inRead invented by Teads
The highly-detailed index, published by data website FiveThirtyEight , rates 636 professional football clubs across world football based on a series of stats and analytics including expected goals for and against, strength of league and performance in the transfer market.

And in the latest table published on Sunday, Rangers are the highest placed Scottish side coming in 92nd - 14 places above Celtic .
That's despite the Hoops' utter dominance of the Scottish league in recent seasons with Neil Lennon looking to clinch a ninth successive title in the upcoming campaign.

Rangers have an offensive rating of 1.9 and a defensive rating of 1.1 (expected goals for and against an average team) according to the stats. That gives Steven Gerrard 's side an SPI (Soccer Power Index) rating of 64.2.
Celtic, in 106th place, have been given an SPI of 62.0 with their offensive rating at 1.6 and defensive 0.9.


Of course the stats mean little ahead of real stuff kicking off on August 3. Leeds United (91st), Burnley (85th) and Bournemouth (79th) all sit above the Old Firm duo while Sheffield United (101st) are higher rated than Celtic.

And you have to drop way below the Lutons, Portland Timbers and FC Tokyos of world football to find the third ranked Scottish side.
That's Kilmarnock who are ranked a lowly 303rd with an offensive rating of 1.0 and defensive rating of 1.1 giving Angelo Alessio's team an SPI of 41.9.
Aberdeen are 343rd with an offensive rating of 1.3 and a defensive one of 1.5 for an SPI of 39.2.


Then come Hibs in 399th, Hearts in 498th, Motherwell 522nd, St Johnstone 569th, St Mirren 571st and Hamilton 618th.
Premier League champions Manchester City top the rankings with an SPI of 94.4 while Bayern Munich sit second (93.5) and European champions Liverpool third (92.9).
Manchester United are way down in 34th place with an SPI of 75.7.
For the full 636-team rankings click here.
 
Last edited:
All over Glasgow...

giphy.gif
 
Morelos2106 said:
I just find it odd that a model has us above Celtic and people are immediately picking holes in it, as if such a thing were impossible.
Click to expand...
I agree i think we are stronger than them now, i would have our squad over theres as it stands
 
538 would be the first to acknowledge that they are not very good at football. Though they’d call it soccer.

I’d also wonder what difference losing Morelos would do to that. We look like we are sitting high because we have very high attacking stats. Some of Alfie’s numbers will be exceptional there.

Nonetheless, nice to see and will really get under Celtic fans’ skins.
 
Starting Anew said:
538 would be the first to acknowledge that they are not very good at football. Though they’d call it soccer.

I’d also wonder what difference losing Morelos would do to that. We look like we are sitting high because we have very high attacking stats. Some of Alfie’s numbers will be exceptional there.

Nonetheless, nice to see and will really get under Celtic fans’ skins.
Click to expand...
It's like any model, the more data they feed into it, the better it will become. Give it a couple of seasons.
 
Bowery Boy said:
Cool so what's it telling us?
Click to expand...
https://fivethirtyeight.com/methodology/how-our-club-soccer-predictions-work/

At the heart of our club soccer forecasts are FiveThirtyEight’s SPI ratings, which are our best estimate of a team’s overall strength. In our system, every team has an offensive rating that represents the number of goals it would be expected to score against an average team on a neutral field, and a defensive rating that represents the number of goals it would be expected to concede. These ratings, in turn, produce an overall SPI rating, which represents the percentage of available points — a win is worth 3 points, a tie worth 1 point, and a loss worth 0 points — the team would be expected to take if that match were played over and over again.

It's broken down further in the article.
 
Morelos2106 said:
I just find it odd that a model has us above Celtic and people are immediately picking holes in it, as if such a thing were impossible.
Click to expand...

I'll be honest mate in that Ive only had a cursory glance at it but is the model not based on what has gone before?

IE we have won no trophies in the past 3 years while the yahoo has won all 9.

So really I dont think these stats based analysis' although interesting to read, actually mean that much until we as a club and team actually go out and win more points than the mentally challenged over a full season and win this league back.

Again only my humble(possibly wrong) opinion but thats how I see things.

Ps just for the record I would rather we were ahead of them in this model than behind them BUT would much rather we were ahead of them in the league come the last day of the season.
 
Morelos2106 said:
https://fivethirtyeight.com/methodology/how-our-club-soccer-predictions-work/

At the heart of our club soccer forecasts are FiveThirtyEight’s SPI ratings, which are our best estimate of a team’s overall strength. In our system, every team has an offensive rating that represents the number of goals it would be expected to score against an average team on a neutral field, and a defensive rating that represents the number of goals it would be expected to concede. These ratings, in turn, produce an overall SPI rating, which represents the percentage of available points — a win is worth 3 points, a tie worth 1 point, and a loss worth 0 points — the team would be expected to take if that match were played over and over again.

It's broken down further in the article.
Click to expand...
Interesting.
 
LoganBear said:
I'll be honest mate in that Ive only had a cursory glance at it but is the model not based on what has gone before?

IE we have won no trophies in the past 3 years while the yahoo has won all 9.

So really I dont think these stats based analysis' although interesting to read, actually mean that much until we as a club and team actually go out and win more points than the mentally challenged over a full season and win this league back.

Again only my humble(possibly wrong) opinion but thats how I see things.

Ps just for the record I would rather we were ahead of them in this model than behind them BUT would much rather we were ahead of them in the league come the last day of the season.
Click to expand...

They don’t forecast reality. The tell you statistical odds. From their measurements, they think that we should beat Celtic. If you ran the next season 1000 times, we’d win it 508 times (I think). According to their model. Reality could very easily be one of the 492 others.

One interesting thing is that part of their model is based on looking at players’ transfermrkt valuations as they think that market value correlated with contribution to teams. I don’t know if they have based that correlation on real transfer prices or on transfermrkt valuations. If the former, this may contribute to their fairly poor outcomes on football as transfermrkt valuations are rubbish. If they have run the correlations based on them then fair enough but if they are assuming that they are an easily accessible way of getting real values then they’ll be in trouble.

This could actually make our lead greater. Our players are hugely and bizarrely undervalued on transfermrkt. No idea how they rate Celtic ones. Aribo for instance is valued at £180k in spite of us probably paying more as a development fee! And Ojo is valued at £1.8mill and you are barking if you think Liverpool would sell for that.
 
LoganBear said:
I'll be honest mate in that Ive only had a cursory glance at it but is the model not based on what has gone before?

IE we have won no trophies in the past 3 years while the yahoo has won all 9.

So really I dont think these stats based analysis' although interesting to read, actually mean that much until we as a club and team actually go out and win more points than the mentally challenged over a full season and win this league back.

Again only my humble(possibly wrong) opinion but thats how I see things.

Ps just for the record I would rather we were ahead of them in this model than behind them BUT would much rather we were ahead of them in the league come the last day of the season.
Click to expand...
Explanation of the model is above. It's been crafted from over 550,000 matches. Which, contrary to what I said to SA earlier, feels like a sufficient dataset for a model.

But bear in mind, these things are not designed to measure absolutes. They measure probability. It's really worth reading the explainer, which is far more informative than I could ever be.
 
Morelos2106 said:
Explanation of the model is above. It's been crafted from over 550,000 matches. Which, contrary to what I said to SA earlier, feels like a sufficient dataset for a model.

But bear in mind, these things are not designed to measure absolutes. They measure probability. It's really worth reading the explainer, which is far more informative than I could ever be.
Click to expand...

I think that the thing that gives them a problem is that football is very low scoring. A single fluke or mistake can change a result. In a higher scoring game (like all the US ones) these random factors level out over a game to an extent and the underlying fundamentals show through. In football freak results are more common making statistical prediction hard.
 
Starting Anew said:
I think that the thing that gives them a problem is that football is very low scoring. A single fluke or mistake can change a result. In a higher scoring game (like all the US ones) these random factors level out over a game to an extent and the underlying fundamentals show through. In football freak results are more common making statistical prediction hard.
Click to expand...
What makes football easier to predict over the long-term is that budgets are relatively stable, and the gap between budgets is relatively stable. In the NFL, teams are terrible one season and great the next. There's always a team that goes from worst-to-first. That's really difficult to predict.

With regards to short-term prediction, i.e. over a single match, you're right, it's harder to predict than most American sports. But to my mind (and I haven't thought a huge amount about it) that would just mean that we have more statisical outliers, e.g. the Everton 4-0 Man City game mentioned in the article. The fundamentals are still likely to be strong if they're based on an xG model, though maybe not as strong as American sports. I'd still be happy to place wagers based on a predictive xG model.
 
Morelos2106 said:
What makes football easier to predict over the long-term is that budgets are relatively stable, and the gap between budgets is relatively stable. In the NFL, teams are terrible one season and great the next. There's always a team that goes from worst-to-first. That's really difficult to predict.

With regards to short-term prediction, i.e. over a single match, you're right, it's harder to predict than most American sports. But to my mind (and I haven't thought a huge amount about it) that would just mean that we have more statisical outliers, e.g. the Everton 4-0 Man City game mentioned in the article. The fundamentals are still likely to be strong if they're based on an xG model, though maybe not as strong as American sports. I'd still be happy to place wagers based on a predictive xG model.
Click to expand...

Yeah. They have tried their best to account for it.

But with leagues often coming down to a couple of games’ points differential then the large spread of probabilities in individual games makes it tricky. Especially in a close league, it won’t be very predictive.

The fundamentals sound pretty good though so seeing us on top is great.

I find their table very hard to understand though - it shows us as the best team but looks like it correctly predicted Celtic winning last year. Is that because of a weird probability spread or have they updated the SCI rankings but not the table?
 
Starting Anew said:
Yeah. They have tried their best to account for it.

But with leagues often coming down to a couple of games’ points differential then the large spread of probabilities in individual games makes it tricky. Especially in a close league, it won’t be very predictive.

The fundamentals sound pretty good though so seeing us on top is great.

I find their table very hard to understand though - it shows us as the best team but looks like it correctly predicted Celtic winning last year. Is that because of a weird probability spread or have they updated the SCI rankings but not the table?
Click to expand...
They haven't updated the table, it's still for last year.

Said before on this forum that luck is a much underrated factor in a league campaign. People don't like to imagine things coming down to sheer chance - it's much easier to pin it on players/managers/referees. To win a league, you have to be lucky as well as good, especially if it's close.

It's why the bookies always give you better value on the unders, instead of the overs.
 
“Utter dominance” you can almost hear his teeth grinding as he batters that into his laptops keyboard :D
 
You must log in or register to reply here.

Share this page

Latest posts

Back
Top